For the first time ever, scientists have mapped an insect’s brain with more than half a million synapses

Scientists have created a 3D reconstruction of the fruit fly’s brain, which contains more than 3,000 neurons and half a million synapses. The new 3D reconstruction reveals the architecture of the insect nervous system, including neuron types and their interconnections, and provides insights into how the brain processes information and generates behaviour. The study shows that the brain is highly interconnected, with abundant feedback from neurons and multiple new circuit motifs.

Some features of the Drosophila brain are similar to advanced deep learning architectures, providing a basis for future studies of neural circuits. The findings pave the way for a better understanding of brain function, with potential applications in artificial intelligence and robotics.

A team of scientists led by Professor Winding of the University of Cambridge used electron microscopy to image the brain of the Drosophila melanogaster larva and accurately reconstruct its synaptic connections. Their findings have been published in the journal Science. Analysis showed that the caterpillar’s brain contains 3,016 neurons and 548,000 synapses. The researchers conducted a detailed analysis of the brain’s circuit architecture, including types of connections and neurons, network nodes, and circuit decorations.

The team also found extensive multisensory integration in the brain and multiple interconnected pathways of varying depth from sensory neurons to efferent neurons that form a distributed processing network. The brain had a highly repetitive structure, with 41% of neurons receiving repeated input over long distances. However, this was not evenly distributed and was particularly high in domains involved in learning and choice of work. Dopaminergic neurons that control learning were among the most frequently replicated neurons in the brain.

Artificial intelligence applications

It is also possible that some of the architectural features observed in the brain of Drosophila larvae can be used to provide information about artificial intelligence (AI) architectures. In particular, the repetitive nature of Drosophila synapses may provide insight into how AI algorithms can be designed to more closely mimic human decision-making processes.

The newly established synaptic connectivity of the Drosophila larval brain provides valuable insights into its functional roles and circuit motifs. It is possible that some of these findings could have applications in artificial intelligence and robotics, which could lead to improved decision-making algorithms and learning processes.